Main

RSV causes significant respiratory disease, infecting almost all children during the first 2 y of life (1, 2). Indeed RSV is the most frequent cause of bronchiolitis and pneumonia in infants requiring hospitalization (3). Moreover, at least 50% of infants who have acute viral bronchiolitis because of RSV have subsequent episodes of wheezing consistent with asthma (4, 5). However, the pathophysiology of RSV bronchiolitis is unclear. RSV bronchiolitis is characterized histologically by epithelial necrosis and peribronchoalveolar infiltration with leukocytes, especially neutrophils (6, 7). We recently reported that neutrophils can augment epithelial leakage and detachment in RSV infection in vitro (8), suggesting that neutrophils and their products may have an important role in the clinical and pathologic changes occurring in RSV infection. However, it is unknown whether such epithelial damage occurs in vivo in infants with RSV bronchiolitis, as there has been no clinical measure available to detect such damage.

The alveolocapillary membrane, with a surface area approximately 50 times that of the body and only 0.1–0.2 μm thick, faces an extraordinary task in partitioning proteins of the plasma and those of the pulmonary ELF (9). We have recently shown that impairment of the alveolocapillary barrier allows increased leakage of SP-A and SP-B into the circulation in a manner that reflects lung function and injury score (9–11). Because the integrity of the bronchoalveolar epithelium is likely to be impaired in RSV bronchiolitis, we have compared the plasma concentrations of SP-A and SP-B in infants without cardiopulmonary diseases with those in infants hospitalized because of RSV bronchiolitis.

Methods

Subjects

Fifteen RSV-positive infants between 1 and 14 mo of age (6.0 ± 1.1 mo; mean ± SEM), admitted for acute bronchiolitis to Flinders Medical Center in Adelaide, South Australia, were enrolled. The RSV antigens were detected from an aliquot of nasopharyngeal aspirates by indirect immunofluorescence assay. The infants, who were healthy until the time of infection, all had a severe infection with coughing, wheezing, chest overinflation, and tachypnea. None of them required intubation or ventilation. A healthy control group of 13 infants between 1 and 24 mo of age (7.7 ± 1.7 mo), who required blood tests before minor surgery, were also enrolled. Informed consent was obtained from the parents of the infants, and the study was approved by the Clinical Investigations (Ethics) Committee at Flinders Medical Center.

Blood sampling and plasma preparation

Blood (1–2 mL) was drawn from an antecubital vein into lithium heparin tubes from patients and control infants. The blood was immediately centrifuged at 913 ×g for 5 min at 4°C (GS-6R Centrifuge, Beckman, USA). The plasma was collected and stored at −20°C or −80°C for batch analysis.

Determination of SP-A and SP-B

Our SP-A and SP-B ELISA have been previously published (9–11). Briefly, to free the SP-A and SP-B from any associated plasma or surfactant components, aliquots are first treated with EDTA, SDS, and Triton X-100. The antigens are measured by ELISA inhibition assays using antibodies Po-A and Po-B raised against alveolar proteinosis-derived SP-A and mature SP-B, respectively (9–11). All samples are assayed in duplicate at four serial dilutions. Standards, assayed in quadruplicate, are included in each ELISA plate at eight serial dilutions (ranging from 7.8 to 1000 ng/mL). Absorbance is measured using a Dynatech MR5000 reader (Dynatech Laboratories; Chantilly, VA). An AssayZap program (Biosoft; Ferguson, MO) is used to generate a standard curve (r> 0.99) and to compute the concentrations of SP-A and SP-B. The antibodies do not react with other plasma or ELF antigens, and the assays have a coefficient of variation of approximately 6%.

Antibody specificity

SP-A.

The specificity of Po-A has been described in detail elsewhere (9). Briefly, its specificity was compared by Western blotting with the MAb Pe-10 (9). The reactivity was tested under both reducing and nonreducing conditions using whole human bronchoalveolar lavage fluid, human surfactant, purified human SP-A, and human serum. The patterns of immunochemical staining obtained were indistinguishable regardless of whether Po-A or PE-10 was used and corresponded solely to immunoreactive SP-A (9).

SP-B.

The specificity of Po-B has also been described in detail elsewhere (10). Briefly, we have purified mature human SP-B and raised a polyclonal antibody that reacts strongly with the mature protein, the processing intermediate, and the SP-B precursor when separated under both reducing and nonreducing conditions. We have compared its specificity with antibodies raised against a synthetic fusion protein of the human full-length SP-B cDNA; antibodies generated against recombinant human C-terminal propeptide, residues 280–381; and finally, antibodies generated against recombinant human N-terminal propeptide, residues 24–200 (10). Our analysis suggests that plasma contains very little mature SP-B and that the major immunoreactive forms present correspond to the precursor (approximately 42 kD) and processing intermediate (approximately 25 kD) (10).

Statistical analysis

Data are given as mean ± SEM. Differences in results between groups were examined with independent-samples t test. A p value of < 0.05 was considered significant.

RESULTS

The plasma concentration of immunoreactive SP-A in the control infants (673 ± 77 ng/mL) was not different from that in the infants with RSV-positive bronchiolitis (630 ± 94 ng/mL;Fig. 1). In contrast, plasma immunoreactive SP-B was significantly higher in the infants with RSV-positive bronchiolitis (4017 ± 852 ng/mL) than that in the controls (1313 ± 104 ng/mL;p< 0.01;Fig. 1). Consistent with this, the SP-B/SP-A ratio in plasma was approximately 300% higher in the infected group. Neither the plasma concentration of immunoreactive SP-B nor the SP-B/SP-A ratio was related to age.

Figure 1
figure 1

Plasma concentrations of SP-A and SP-B in control infants (n= 13) and infants with RSV-positive bronchiolitis (n= 15). Both SP-A and SP-B were determined using ELISA. The data are expressed as the mean ± SEM. *p< 0.05 compared with control group.

Full size image

DISCUSSION

Our data illustrate that the plasma SP-B, but not the SP-A, concentration is elevated in RSV-positive infants with bronchiolitis. We believe that the site of leakage is the alveolocapillary barrier, inasmuch as the bronchocapillary barrier has a surface area approximately 2.5% to 5% that of the alveolocapillary barrier and is considerably thicker (several micrometers to millimeters as opposed to 0.1 to 0.2 μm). In addition, protein transport across the bronchial epithelium is usually accompanied by degradation, whereas proteins cross the alveolar epithelium largely intact (12).

Consistent with this, there is now a large body of work describing the flux of locally secreted proteins from the hypophase into the circulation (13). Moreover, because of difficulties associated with the bioavailability of orally ingested recombinant therapeutic proteins, increasing attention has been given to the use of the lung as a noninvasive means of systemic delivery. Nebulized intrapulmonary recombinant human insulin has been used (14), and recombinant human superoxide dismutase administered intratracheally to infants with respiratory distress syndrome is found in significant quantities in serum and urine within 6 h of administration (15).

In the airspaces, SP-A predominantly forms high-molecular weight oligomers (approximately 650 kD) with Stokes radii of approximately 35 nm (16). Although mature SP-B, which associates as a low Mr (approximately 18 kD) thiol-dependent homodimer (17), is normally intimately associated with complexes of surfactant phospholipids, which are possibly too large to readily breach the alveolocapillary barrier, we have previously shown that at least some of the protein is secreted into the alveolus as hydrophilic, monomeric proprotein and processing intermediate with approximate Mr of 45 kD and 25 kD, respectively (10). We have shown that these forms breach the alveolocapillary barrier more readily than SP-A (10, 11), presumably because they are smaller, and so provide a better marker of lung injury. Therefore, our finding that RSV bronchiolitis is associated with increased levels of SP-B, but not SP-A, may reflect the relative severity of the injury, the nature of the sieving properties of the lung, and the differences in the sizes of the proteins (9–11). This indicates that RSV bronchiolitis requiring hospitalization is associated with increased alveolocapillary permeability, but, at least in cases in which ventilatory support is not required, to a lesser extent to that which occurs in acute respiratory distress syndrome (9, 10). As far as we know, this is the first report identifying increased alveolocapillary permeability in infants with RSV bronchiolitis.

Consistent with the notion that our findings reflect increased alveolocapillary permeability and the difference in the sizes of SP-A and SP-B, Kerr and Paton (18) recently found that SP-A and SP-B are reduced approximately 40% and 90%, respectively, in bronchoalveolar lavage fluid harvested from children infected with RSV. This indicates that neither selective degradation nor consumption of alveolar SP-A, or increased SP-B synthesis, are likely to explain why plasma SP-B is elevated in our RSV-infected cohort, whereas SP-A is not. Indeed, increased leakage of SP-B may be a contributing factor to the decreased SP-B concentrations with regard to those of SP-A in bronchoalveolar lavage fluid harvested from children infected with RSV (18).

The mechanism of lung injury induced by RSV infection in vivo is unknown. In most cases, a common pathway involving the recruitment and activation of leukocytes, and in particular neutrophils, is thought to result in diffuse inflammatory lung injury (10). Neutrophils play an important role in lung injury in acute respiratory distress syndrome. They are also the predominant inflammatory cells in RSV infection (6, 7), and our previous in vitro study has shown that RSV infection can induce alveolar type II-like cell (A549) leakage and detachment to a certain extent, both of which are augmented by neutrophils (8). Therefore, we speculate that alveolar permeability may be increased both through the direct action of the virus, as well as through that of the neutrophils. In addition, increased work of breathing associated with airway obstruction can generate large negative alveolar pressure and cause edema (19, 20), whereas lung hyperinflation is a well-known factor increasing alveolocapillary permeability (21). However, we do not know the relative contribution of these factors to the overall alveolocapillary permeability.

Pulmonary surfactant is known to stabilize alveoli and to maintain patency of conducting airways (22–24). In a guinea-pig asthma model challenged with ovalbumin aerosol, it was shown that surfactant dysfunction developed, which might cause conducting airways to become blocked by liquid columns and thereby increase airway resistance (24). It has also been reported that RSV infection in BALB/c mice induced surfactant dysfunction, which may increase airway resistance (25). It seems that surfactant may play an important role in the pathophysiology of both RSV infection and asthma (25, 26). It has recently been shown that surfactant from infants with viral bronchiolitis has an increased minimum surface tension (27). Inasmuch as SP-B is essential for normal surface activity (28), possible leakage of SP-B into the circulation may reduce alveolar levels and contribute to the reduced surface activity observed by Dargaville and associates (27).

In summary, our study suggests that alveolocapillary permeability is increased in infants with RSV bronchiolitis in vivo. However, currently we do not know whether the circulating concentration of SP-B is related to disease severity. Further studies are required to elucidate the role of surfactant and pulmonary injury in the pathophysiology of RSV bronchiolitis.